Isolation and analysis of NBS-LRR analogs in Cerasus clarofolia

Abstract:【Objective】A line of Cerasus clarofolia Schneid originated from Moutain Leigong (GuizhouProvince, P. R. of China) demonstrates high resistance to crown gall of cherry. In order to understandthe resistance mechanism to this disease, as well as to clone the candidate resistance gene, the NBS-LRR（nucleotide binding site-leucine rich repeat）analogs were cloned, and the evolution pattern was furtheranalyzed.【Methods】Degeneracy primers based on the conserved domain of NBS-LRR were used toisolate NBS-LRR sequences from genomic DNA of C. clarofolia. An approximate 500 bp fragment wasobtained by polymerase chain reaction (PCR), and then the aplicons were cloned, sequenced and analyzed.The inserted fragment in the pGEM-T easy vectors sequences was isolated using VecScreen onlinetools. Nucleotide acid and deduced amino acid sequences were analyzed via BLASTn and BLASTx(https://blast.ncbi.nlm.nih.gov/Blast.cgi). The length of nucleotide sequence and AT/CG proportion were estimated using DNA star software. Multiple sequences alignment was generated by Clustal Wand visualized using BoxShade. The phylogenetic tree from nucleotide acid and deduced amino acid sequenceswas constructed using Mega 6.0 program. The nonsynonymous substitutions (Ka) and synonymoussubstitutions (Ks) were estimated using Dnasp5.0.【Results】A total of 36 unique clones fromNBS-LRR resistance gene analog (RGAs) were isolated, which were named as PpRGA1-PpRGA36.The sequences were submitted to the GenBank with the accessions as ACK102301-ACK102336. Nucleotidesequences ranged in length from 429 bp to 516 bp, and homology of the 36 sequences varied inrange of 41.7% to 99.8% and 19.0% to 100.0% among nucleotide sequences and amino acid sequences,respectively, revealing rich diversity of NBS-LRR in this species. Three sequences contained deletionmutation, frameshift mutation or premature stop codons. AT/GC ratio was in range of 1.37 (PpRGA18)-1.58 (PpRGA10). The phylogenetic tree generated from nucleotide acid sequences might divide NBSLRRinto four clades (Ⅰ- Ⅳ), clades Ⅰ- Ⅲ were classified as TIR type, and clade Ⅳ was classified asnon-TIR type. As multiple amino acid sequences alignment, the sequences contained conserved domainmotifs such as P-Loop, Kinane-2 and GLPL. According to the characteristics of Kinane-2 motif, 24 sequenceswere grouped into non-TIR, and 12 into TIR type. To learn more about the evolutionary relationshipsamong PpRGAs, a total of 144 NBS-LRRs from GeneBank, whose homology was above 70%with each subclade of PpRGAs, were used to construct the phylogenetic tree with 36 PpRGAs. The resultshowed that 40 and 104 NBS-LRRs clustered with non- TIR and TIR, respectively. Of the non-TIRs, all the NBS-LRRs were from Rosaceae species such as P. persica, P. mume, C. avium, P. bretscheideri,Fragaria × ananassa, Malus domestica, Rubus galanucus. Of the TIRs, a total of 93 NBSLRRsfrom Rosaceae, and 11 NBS-LRRs from other families such as Juglandaceae, Leguminosae, Cucurbitaceaeand Fagaceae. Among PpRGA specific clades, 17 P. persica, 13 P. mume, 5 P. avium and 1P. kansuensis NBS-LRRs were grouped with 24 PpRGAs, indicating that C. clarofolia had more closeevolutionary distance with P. persica, P. mume than P. avium, and those clades showed Prunus L. species-specific pattern, inferring the genes have undergone gene expansion after speciation of Prunus L.By considering the category of species clustered result among TIR and non-TIR type, we inferred thatTIRs were more ancient than non- TIRs. Furthermore, by estimating average of Ka/Ks ratios amongeach clades, clade Ⅳ (belong to non-TIR) was greater than 1, cladeⅠand Ⅲ were less than 1 and cladeⅡ was approximately equal to 1, suggesting that there was an evolving process under a neofunctionalization,subfunctionalization and nonfunctionalization model driven by positive selection, purificationselection and neutral selection, respectively. In addition, the distribution range of Ks values of TIR andnon-TIR was calculated, and interestingly, TIR distributed in two regions, which ranged in 0-0.3 and0.8-12, and non-TIR distributed in range of 0-0.3, suggesting that duplication events of TIR occurredearlier than non-TIR. The frequency of Ks values commonly used to measure duplication scale, by calculationof obtained PpRGAs and TIR, was highly greater than that of non-TIR, indicating that TIR hadmore duplication events than non-TIR. In addition, 81% of Ks value frequency distributed in range of 0-0.3, and 64% of Ks frequency of non-TIR was in range of 0-0.1, suggesting that large-scale duplicationoccurred both in non-TIR and TIR in recent period. Finally, the relationship between Ks value and Ka/Ks was analyzed, and in each Ks value distribution region, the Ka/Ks value got smaller as Ks value increased.【Conclusion】Three clades of TIR and one clade of non-TIR were found in C. clarofolia NBSLRRs.The phylogenetic tree as well as homology and Ks value analysis showed TIR may originatefrom more ancient time than non-TIR, TIR of C. clarofolia got revolution relationship with P. persica,and P. mume was more closely related than C. avium. Both TIR and non-TIR had relatively large-scale duplication events that occurred in recent period. Ka/Ks value inferred that different selective pressureand duplication pattern existed between each clade of non-TIR and TIR, suggesting that those mighthave different evolutionary pattern during long-term history.